Multi-Guide Set-Based Particle Swarm Optimization for Multi-Objective Portfolio Optimization

Portfolio optimization is a multi-objective problem (MOOP) with risk and profit, or some form of the two, as competing objectives. Single-objective portfolio requires trade-off coefficient to be specified in order balance two Erwin Engelbrecht proposed set-based approach single-objective optimization, namely, particle swarm (SBPSO). SBPSO selects sub-set assets that search space for secondary task optimize asset weights. The authors found was able identify good solutions problems noted benefits redefining problem. This paper proposes first (MOO) SBPSO, its performance investigated optimization. Alongside this investigation, multi-guide (MGPSO) evaluated compared against algorithms. It shown competitive algorithms, albeit multiple runs. i.e., (MGSBPSO), performs similarly other algorithms while obtaining more diverse set optimal solutions.